Editorial & Testing Policy
At Zynocode Labs, we believe online utilities and technical resources must be demonstrably accurate, strictly private, and verified against reproducible tests. This policy outlines our engineering standards, our client-side runtime boundaries, and how we test every calculation and guide before publication.
1. Browser-Side Execution & Zero File Uploads
Unlike traditional cloud convertors that transmit sensitive documents to remote servers, Zynocode Labs tools run entirely inside your browser sandbox. Your images, PDFs, source code, text entries, and financial numbers never leave your device.
Native Web Platform APIs We Rely On:
- PDF Processing: Embedded Mozilla PDF.js (
pdfjs-dist) andpdf-librunning in your browser, with the PDF.js worker served from this site (no third-party CDN). - Image Processing: Native HTML5 Canvas API, OffscreenCanvas, and modern WebP / JPEG encoders.
- Cryptography & Passwords: Native
window.crypto.getRandomValuesand the bcryptjs library (plain JavaScript, run in a Web Worker). - Audio & Speech: The browser's built-in Web Speech API.
- AI tools (background remover, text summarizer, text humanizer): open-weight models run in your browser with Transformers.js. The model files are downloaded once from Hugging Face and cached; your images and text are processed on your device.
Honest Scope of Server Routes (/api/*):
The tools on these pages do not send your files or text to our servers. Our server routes handle sign-in sessions (NextAuth), workspace management, newsletter signups, contact messages, and usage logs for signed-in accounts only (tool name, duration, status and settings such as the chosen format — never your file or text contents; anonymous visits are not logged).
Separately, developers can call the public Tool API (/api/v2/tools, used by the @zynocode/sdk package) with an API key. That API processes the data the developer sends to it on our server; the website's tools do not use it.
2. Automated Test Suites & Real-Output Verification
We do not rely on mock claims. Every tool and content page is verified through automated headless browser test suites located in our scripts/ directory prior to release:
UI & Edge Case Suites
Automated regression suites (test-batch3.cjs through test-batch7.cjs) verify real DOM interactions, input sanitization, copy-to-clipboard, zero horizontal overflow on 375px mobile viewports, and local web worker isolation.
Exact Target Size Verification
Our image and PDF compression test runners (test-compress-to-size.cjs and batch tests) test real binary files against multi-stage compression ladders. A result is only marked valid if the final byte size is strictly within the target threshold and never larger than the original.
Engine Tool Contracts
Our core pipelines in scripts/test-engine-tools.ts validate input schemas, error boundaries, and output contracts across all text, developer, and data utilities.
Recomputed Worked Examples
Via scripts/verify-finance-content.cjs, every single numeric claim and worked example across our finance tool pages is programmatically generated and verified by executing the tool's own calculation functions.
3. Authoring, Sourcing & Citations
All technical guides and documentation on Zynocode Labs are written and reviewed by our Founder & Technology Strategist, Prashant Sharma. We adhere to strict factual guidelines:
- Official Government & Regulatory Sources: Form upload rules, dimension limits, and file size constraints are verified against official recruitment notices (such as the SSC CGL 2024 notice, UPSC Civil Services guidelines, and IBPS exam guidelines). Tax formulas reflect the latest GST Council rates and Income Tax rules.
- Authoritative Technical Documentation: Web performance and SEO guidelines are verified against Google Search Central documentation, Chromium source specifications, and W3C web standards.
- No Fabricated Data: We prohibit speculative statistics, unverified benchmark claims, and vague "studies show" assertions. If a metric cannot be directly cited or tested, it is not published.
4. Date-Stamped Updates & Error Reporting
Technology and regulatory norms evolve constantly. To maintain transparency, every guide and tool page maintains public timestamps:
Both published dates (datePublished) and last-modified dates (dateModified) are recorded in the article byline and rendered into Schema.org Article structured data. When official rules change or tools undergo substantial algorithmic updates, articles are revisited and updated accordingly.
Found a Discrepancy or Have Feedback?
We welcome bug reports, formula corrections, and suggestions from developers and researchers.
Meet the Author
Prashant Sharma
Founder & Technology Strategist, Zynocode IT Solutions